Tao Zhang

This paper proposes a comprehensive taxonomy (SLOT) to systematically categorize security risks, attacks, and defenses specific to Retrieval-Augmented Generation (RAG), clarifying that these risks are distinct from inherent LLM flaws.

The paper introduces DEJA, an automated black-box attack framework that generates stealthy adversarial documents to induce 'soft failures' in RAG systems, degrading utility without triggering overt refusals.

TRACER introduces a novel turn-level reinforcement framework that enables cooperative multi-LLM reasoning by separating decision-making into a regret-matching controller and a generation-credit layer.

The paper introduces StreamSynth, a sequential setting for synthetic data generation, and proposes SynLearner, a framework that enables LLMs to improve synthesis performance by accumulating and transferring experience across a stream of tasks.

The paper introduces Source-Grounded Semantic Reinforcement Learning (SG-SRL), a framework that leverages abundant source-language monolingual data to improve target-language generation in low-resource settings by providing cross-lingual semantic supervision.

The paper introduces DistractionIF, a benchmark showing that larger LLMs are paradoxically less robust to benign, instruction-like noise in reference text, suggesting reinforcement learning can restore this robustness.

The paper proposes EAGLE, a novel evidence-aligned multi-agent framework, demonstrating that requiring shared visual evidence among agents is crucial for achieving reliable and trustworthy consensus in multimodal Visual Question Answering (VQA).

The paper introduces Andes, a framework that treats data generation as a plug-and-play agent skill, enabling autonomous alignment of LLMs by providing an intelligent, closed-loop data synthesis interface.

This paper proposes ShopX, a model-centric framework for intent-driven shopping experiences using a single foundation model for intent understanding, execution planning, and item-space operations.

The paper proposes Fuzzy-Function Programming and introduces Program-as-Weights (PAW), a compact, locally-executable neural artifact for everyday programming tasks.

The paper introduces WaspMOT, a new benchmark for long-term multi-object tracking, and evaluates five tracking-by-detection methods.

JoyNexus is a multi-tenant service for VLA model supervised fine-tuning, reinforcement learning, and evaluation, which decouples services, introduces group batching, and improves training efficiency.